FabCon and SQLCon 2026 in Barcelona: Building the data foundation for Microsoft Copilot and agents

One of the most remarkable aspects of the Fabric and SQL communities lies in the eagerness of individuals to share their insights and discoveries. This week, close to 5,000 data specialists are gathering in Barcelona to exchange knowledge. Over more than 130 sessions and 10 tutorials, along with community meetups and discussions, attendees will share valuable insights and hands-on experiences that can be implemented in various organisations.

The enthusiasm in Barcelona reflects the growing momentum of Microsoft’s data portfolio. Microsoft Fabric has now gained the trust of over 40,000 customers globally. For decades, Microsoft SQL Server has supported enterprise workloads, powering more than 90% of Fortune 500 companies. Prestigious customers like Levi Strauss, H&M, and Thomson Reuters rely on these tools to integrate their data, contextualise AI within their business, and scale operations while ensuring security and governance.

Today, we are thrilled to unveil new advancements in Microsoft Fabric and our database offerings, aimed at aiding organisations in their Frontier Transformation. Notable innovations include seamless integrations with the updated Microsoft Copilot, the ability for agentic app development within Microsoft Power BI, comprehensive observability features in Fabric, our database hub, and much more.

Driving Frontier Transformation with Microsoft Copilot and Microsoft Fabric

The landscape of software development has transformed significantly. AI is evolving, assisting developers not only in writing code quickly but also in managing the entire development phase—from planning and coding to testing and deployment.

Now, the goal is to extend this transformation throughout the entire business. However, companies require more than just models and agents; they need reliable context that includes governed data, shared definitions, operational insights, and frameworks that empower AI to operate reliably.

This serves as the cornerstone of Frontier Transformation: integrating AI into essential business processes, anchored in an organisation’s specific data, knowledge, and workflows. Microsoft is fostering this evolution through Microsoft Copilot and Microsoft IQ, linking AI with organisational knowledge, data, applications, and business context.

At the heart of this system lies Fabric IQ, an intelligence layer that connects enterprise data, business metrics, and operational context throughout the Microsoft ecosystem. By merging unified data from Microsoft OneLake, validated metrics from Power BI semantic models, and contextual data from ontologies and real-time intelligence, Fabric IQ equips users and agents with a comprehensive understanding of the business landscape.

Empowering the new Microsoft Copilot with trusted data from Fabric IQ

Millions of Microsoft 365 users count on Microsoft Copilot daily for tasks like information retrieval, content creation, data analysis, and taking actions seamlessly within their workflow. With the release of the new Copilot last week, it’s evolving into a cohesive experience where users can assist, delegate tasks, and even develop new solutions. The Home feature merges Chat and Cowork, while the Code aspect allows everyone to create custom solutions. As Copilot undertakes more complex tasks, access to reliable business context becomes increasingly vital.

Today, we are excited to announce that we’re integrating business context from Fabric IQ into Microsoft Copilot, anchoring your work in a shared understanding of the organisation. Users can now access AI-driven insights from Fabric within their Copilot experience, with no extra AI token costs involved. Fabric IQ in Copilot Chat and Cowork is now live, and integration into the new Code experience is on the horizon through the Frontier programme.

For Power BI users, this semantic framework is already operational. Reliable business definitions, key metrics, and organisational insights captured in trusted Power BI semantic models and reports can now directly influence decisions and actions within Microsoft Copilot. This means Copilot users can pose business-related inquiries and receive AI-generated insights rooted in the semantic data layer that their organisation already trusts.

As highlighted by Bogdan Crivat in his Bringing governed analytics into the flow of work: Fabric Analytics at FabCon Europe 2026 blog, this represents a significant step towards conversational analytics that merges enterprise data, business context, and AI to assist users in making informed decisions.

Check out the following demo to see this feature in action:

The Next Step for Power BI: Agentic App Creation

With over 425,000 customers spanning various sectors and countries, Power BI has empowered countless users to transform their data into actionable insights. Today, we are thrilled to unveil the latest advancement in Power BI: introducing agentic app creation within Microsoft Power BI Desktop.

Traditional business intelligence has typically involved analysts modelling data, constructing reports, and publishing dashboards for others to access. With Copilot in Power BI, this process is becoming more conversational, facilitating user exploration of data and report generation through natural language. AI now enables users to extend beyond mere insight discovery, empowering them to create tools, workflows, and applications directly from their data.

The new agentic app creation feature in Power BI Desktop elevates this process further. Users start with a trusted semantic model, articulate the application they need, generate it using natural language, preview and modify it in real time, before publishing it to Microsoft Fabric—eliminating the need to begin with a blank slate.

While Copilot in Power BI aids in data analysis and report creation, this new functionality allows for the creation of tailored data applications. These applications can gather inputs, write back data, preserve shared states, and facilitate operational workflows.

This enhanced experience leverages our new Fabric Apps capabilities, equipping each application with SQL database integration in Fabric, as well as robust authentication and security features. As an application evolves, users can enhance its capabilities and seamlessly transition it to developers and IT teams without the need for a complete rebuild.

Watch the demo below to see the new agentic app creation experience in action:

Crucially, this feature won’t be limited to customers with Microsoft Fabric capacity. Power BI Pro and Power BI Premium Per User customers can access apps in Power BI during preview in the coming weeks, along with Microsoft Fabric customers. This offers Power BI Pro and Premium Per User users access to Fabric Apps and Fabric Database capabilities—up to 1 GB per app—complemented with enterprise-grade security and scalability for hosting. Fabric clients can then extend these apps using other Fabric workloads, enriching their applications with existing data and services.

To delve deeper into this feature and discover other innovations on the horizon for Power BI, check out Mohammad Ali’s blog, Power BI’s Next Chapter: The Evolution of Business Intelligence.

Accelerate Your Path from Idea to Production with Fabric Apps

While AI streamlines prototyping, transitioning an application to production still necessitates teams to connect data, implement business logic, manage infrastructure, and uphold governance. Fabric Apps bridges this gap by providing developers with a structured application layer on Microsoft Fabric.

Since its initial launch in June 2026, thousands of developers have harnessed Fabric Apps to create AI-driven business applications, streamline operational workflows, and design custom experiences based on governed data. Utilizing the open-source Rayfin SDK and CLI, developers can construct application backends and deploy them to Fabric as Fabric Apps.

Recent enhancements allow developers to link applications directly to Fabric warehouses, SQL databases, lakehouses, and Power BI semantic models without duplicating data, all while maintaining existing permissions and business logic. New TypeScript backend functions, a Secret Store, PostgreSQL compatibility, private-by-default access, and integrated usage and performance metrics offer the necessary services to move applications into production.

In combination, these features provide developers with a faster route from concept to secure, governed, and production-ready Fabric Apps, enabling them to spend more time on building unique application experiences rather than piecing together infrastructure.

Explore what you can create with Rayfin:

To learn more about Rayfin, read Sachin Patney’s blog, From Prompt to Production: What’s New in Fabric Apps.

Establishing an AI-Ready Data Foundation with Microsoft Fabric

With Fabric IQ bringing contextual business intelligence into Copilot, applications, and agents, this intelligence must first be created and sustained. Businesses need a robust foundation capable of uniting data across diverse sources, preparing it for analysis, and capturing its business relevance through semantic models and ontologies. The latest Microsoft Fabric advancements simplify the journey of building this foundation from start to finish.

Unify Your Data Estate Using Microsoft OneLake

As organisations scale their AI efforts, merely connecting data isn’t sufficient. Copilots and agents require business context, such as terminology, relationships, instructions, and domain knowledge, to provide accurate responses and make informed decisions. Microsoft OneLake offers a versatile foundation for linking data across clouds, applications, and platforms through mirroring and shortcuts.

Share Governed Data and Context with IQ Sharing

Today marks the launch of the preview for IQ sharing, enhancing the secure sharing of governed data and business context across teams, customers, partners, and ecosystems, extending intelligence beyond company borders. Initially, IQ sharing will support governed tables and files, alongside Markdown agent instructions and RDF ontologies, with plans for native support for Fabric IQ ontologies and Power BI semantic models in the future.

Companies like EY, S&P Global, TomTom, Financial Fabric, and xMentium are already exploring integrations with IQ sharing. For instance, TomTom is incorporating its geospatial and transportation intelligence into Fabric, with data products available via the OneLake catalog.

Together, mirroring, shortcuts, and IQ sharing create an open data network that enables organisations to connect, access, and collaborate on governed data across teams, partners, and ecosystems. This expands OneLake beyond simple data integration to facilitate reusable data products that propel analytics and AI.

Expanding the OneLake Ecosystem

We’re broadening the scope of this open data network through the preview of bidirectional integration with Salesforce Data Cloud 360, the general availability of Google BigQuery mirroring, and closer integration with Microsoft Dynamics 365 Business Central. This allows Salesforce data to become part of the unified data estate without needing to move or duplicate it, ensuring it’s accessible across Fabric for analytics and AI applications, including as business context for Fabric data agents.

Additionally, we’re swiftly expanding the OneLake ecosystem with Atlan, dbt, and ClickHouse, making it simpler to imbue OneLake data with trusted business context, transform it using familiar developer workflows, and engage with it from external analytical engines.

Collectively, these innovations simplify cross-platform, multi-tool, and cloud collaboration without the need for data duplication or complex integrations. To find out more about the exciting innovations coming to Microsoft OneLake, read Dipti Borkar’s blog, FabCon and SQLCon Barcelona 2026: What’s New in Microsoft OneLake and Its Rapidly Growing Ecosystem.

Efficiently Process and Harmonise Data

Linking data is merely the first hurdle. Organisations need to prepare, transform, and operationalise data on a grand scale. To accelerate the journey from data to insights and actions, we’re introducing a data engineering agent in Microsoft Fabric, along with new Copilot features in Fabric Real-Time Intelligence.

Delegate Complex Data Engineering Tasks to Data Engineering Agents

As AI becomes an integral part of the development workflow, data engineers are transitioning from performing every task themselves to orchestrating intelligent systems. Fabric is advancing this shift with the data engineering agent, built using technology from Microsoft’s acquisition of Osmos earlier this year. Currently in preview, the data engineering agent provides autonomous execution for intricate, long-term projects like data platform migrations, lakehouse modernisation, performance enhancements, ETL processes, and data preparation.

Engineers now define the desired results and establish parameters. Within these guidelines, the agent autonomously plans and executes the tasks, validating outcomes and iterating on its approach, enabling teams to progress with complex data objectives while ensuring engineers retain control.

The Fabric data engineering agent is also designed for collaborative efforts spanning individuals and teams. Tasks inherit permissions from the workspace, allowing authorized members to track progress, contribute context, make decisions, and work jointly on the same assignment.

Furthermore, since data engineers operate across numerous development environments, these functionalities extend throughout Fabric and developer tools, including GitHub Copilot, Microsoft Visual Studio Code, Codex, and Claude Code. Engineers can initiate, oversee, and direct the same tasks using their familiar tools without losing context or progress.

Discover more about the Fabric data engineering agent in Bogdan Crivat’s blog, Bringing Governed Analytics into the Flow of Work: Fabric Analytics at FabCon Europe 2026.

Simplify Engagement with Real-Time Data via Copilot

Businesses increasingly seek to act upon live operational signals rather than just historical data reports. New capabilities in Copilot enhance Real-Time Intelligence, enabling teams to comprehend, monitor, and manage real-time data using natural language. Users can now construct Real-Time Dashboards by articulating what they wish to understand, as Copilot assists with event house management and administrative tasks like creating tables, configuring policies, managing schemas, and inspecting cluster statuses.

New schema suggestions and optimisation features further enable organisations to boost performance and cut costs, while the operations agent hastens troubleshooting with AI-based root cause analysis and actionable recommendations. For further insights into Real-Time Intelligence innovations, read Yitzhak Kesselman’s blog, Trusted AI Starts with Microsoft Fabric, Real-Time Intelligence, and IQ.

Curate Context through Fabric IQ

With data unified and well-prepared, organisations must capture the business significance, relationships, rules, and operational context essential for AI. Fabric IQ offers a shared intelligence layer, extending governed business context across Microsoft Copilot, Microsoft Foundry, GitHub Copilot, and custom applications, enabling agents to reason with a common understanding of the business framework.

Speed Up Ontology Development and Governance

Today, we are simplifying the process of defining and overseeing this context layer. We are launching a new agentic experience for creating ontologies from semantic models, OneLake data, and other contextual documents, reducing the effort needed to encapsulate business concepts and relationships.

Rules can now be defined in natural language, linked to ontology entities, properties, and relationships, and integrated into the shared context that personnel and agents use to navigate business operations consistently. Additionally, Power BI measures can be included in your ontology, enabling you to reuse established metrics.

We’re also reaffirming our support for Apache Ossie (currently incubating), a community-driven initiative to set up a vendor-neutral standard for exchanging semantic metadata across analytics, AI, and BI platforms. Collaborating with Snowflake and the wider industry, we aim to make semantic context open and usable across platforms.

Collectively, these abilities enhance the maintenance of shared business meaning and empower agents to navigate governed data with improved accuracy and consistency. Explore more about Fabric IQ in Yitzhak Kesselman’s blog, Trusted AI Starts with Microsoft Fabric, Real-Time Intelligence, and IQ.

Develop and Manage Agents in Fabric

The true potential of a shared intelligence foundation is realised when agents can leverage that context to deliver tangible outcomes. Recent enhancements across Fabric agents simplify the processes of building, governing, and operationalising AI experiences across Microsoft Fabric and other AI platforms.

We’re enhancing the synergy between Fabric data agents and ontologies within Fabric IQ, enabling more accurate and comprehensible answers rooted in business concepts and relationships. The Build Agent with AI mode can now utilize semantic models, documentation, glossaries, and data dictionaries to create richer user experiences with less manual configuration. The Operations Agents are enhancing their functions by adding pre-approved actions and new root-cause analysis features, assisting teams in investigating issues, identifying likely causes, and automating repeatable responses while retaining human oversight. For more details regarding these updates, check out Bogdan Crivat’s blog, Bringing Governed Analytics into the Flow of Work: Fabric Analytics at FabCon Europe 2026, and Yitzhak Kesselman’s blog, Trusted AI Starts with Microsoft Fabric, Real-Time Intelligence, and IQ.

Manage Fabric with Enhanced Visibility and Control

As organisations increasingly rely on a unified data platform, operational maturity becomes as crucial as analytical capability. Teams need to grasp what’s happening across jobs, capacities, workspaces, and services—and resolve any issues before they disrupt the business.

Presenting New Observability Features in Microsoft Fabric

New observability features in Fabric, now in preview, are designed to elevate organisations from reactive troubleshooting to proactive operations. Covering workspace monitoring, the monitor hub, and operations agents, these observability enhancements offer a holistic view of the Fabric ecosystem, simplifying the early detection of issues, AI-assisted root cause investigation, and streamlined management from a unified experience.

The updated workspace monitoring system allows you to group telemetry in a single eventhouse for various workspaces, easing cross-workspace oversight. You can even make this data available across other destinations with custom endpoints while determining the exact retention rates and caching policies you require.

We’ve also improved the monitor hub, providing deeper insights into jobs, capacities, and workloads through built-in views, alerts, and AI-derived insights. New Activator-driven job notifications keep teams informed about job statuses, while a new capacity overview assists in tracking consumption and available resources. Additionally, new investigative functions for operations agents enable data professionals and administrators to swiftly diagnose issues by autonomously identifying root causes and suggesting actions.

Together, these features significantly reduce the effort needed to pinpoint problems, enhancing reliability, and ensuring that data and AI workloads run smoothly.

Check out the demo below to witness observability in action:

Pay Only for What You Use with On-Demand Billing

Visibility is merely one aspect of the equation. Organisations also need flexible, predictable methods to align infrastructure costs with usage as data and AI workloads expand.

Building upon the Autoscale Billing for Spark, on-demand billing broadens consumption-based pricing across Fabric workloads, allowing you to pay only for the compute resources you actually use, automatically adjusting resources up or down based on demand. The Fabric zero-provisioned feature complements on-demand billing by permitting you to embark on using Fabric and OneLake without needing a prior commitment to capacity.

To learn more about observability in Fabric, on-demand billing, and other cutting-edge innovations in the Fabric platform, read Kim Manis’s blog, Build, Deploy, and Govern Microsoft Fabric at Scale.

Manage, Optimise, and Scale Your Database Estate

Building intelligent applications demands more than just high-performing AI models. Organisations also need databases that can scale up with demand, tools to manage their expanding data estates, and intelligent operations to keep their systems running smoothly.

Today, we’re unveiling new features across the Database Hub, Database Agents, and Microsoft Azure SQL Database that enable organisations to build, manage, and optimise the data foundation behind the next generation of intelligent applications.

Centralise Your Database Management with the Database Hub in Fabric and Database Agents

As developers create more applications and as operational workloads shift into Fabric, managing an expanding database estate becomes just as vital as application development. Organisations require a cohesive method to monitor, govern, and optimise these databases.

Building on the Database Hub’s introduction at FabCon & SQLCon Atlanta, it is now in preview. This feature offers a central control plane for managing SQL Server, Microsoft Azure SQL, PostgreSQL, Azure Cosmos DB, and SQL databases within Fabric. From a unified dashboard, administrators can track health, identify risks, enforce governance, and gain visibility across their complete database ecosystem.

We’re also rolling out database agents for SQL and PostgreSQL on Azure, which will soon be available in preview. These AI-driven specialist agents, accessible within the Database Hub in Fabric and VS Code, continuously monitor and analyse database signals, assisting administrators in troubleshooting issues, understanding performance limitations, identifying optimisation chances, and hastening resolution efforts. By integrating AI-supported operations within a unified management interface, teams can transition from insights to action more swiftly while bolstering reliability, performance, and security.

These enhancements collectively provide an operational foundation for intelligent applications throughout Microsoft, helping organisations manage their data estate with the same simplicity, scalability, and intelligence that fuel analytics and AI in Microsoft Fabric.

Learn more about the Database Hub and database agents in Shireesh Thota’s blog, SQLCon Barcelona 2026: Advancing SQL with Greater Control, Scale, and Intelligence.

Enhancing Scale, Efficiency, and AI Innovation in Azure SQL Database Hyperscale

Intelligent applications impose fresh demands on databases, necessitating increased scale, efficiency, and built-in support for AI applications. We are introducing new innovations in Azure SQL Database Hyperscale that help developers minimise costs, scale demanding applications, and create AI-driven experiences using familiar SQL tools.

Serverless auto-pause for Azure SQL Database Hyperscale, currently in preview, automatically pauses compute resources when inactive and reactivates them when workloads resume, aiding cost reduction for development, testing, and intermittent tasks. We’re also raising elastic pool density to support up to 50 databases per pool, with 192 vCore compute for Azure SQL Database Hyperscale now generally available, enabling greater scale and concurrency for mission-critical applications.

Additionally, we’re announcing the general availability of DiskANN vector indexes for Azure SQL Database Hyperscale, Azure SQL Managed Instance, and SQL databases within Fabric, introducing high-performance vector search directly into the database engine. This allows developers to build recommendation engines, conduct semantic searches, and create retrieval-augmented generation (RAG) applications without needing to transfer data to a separate vector store.

Facilitate Sovereign Data across Customer-Controlled Environments

Businesses require the flexibility to deploy data platforms where performance, regulatory, and operational needs dictate. Microsoft is broadening the availability of Azure and SQL Server capabilities, enabling modern workloads across sovereign, customer-controlled data estates, whether in the cloud, on the edge, or in disconnected situations.

SQL Server on Azure Local is now generally available, enabling SQL Server and Azure capabilities in local settings. This allows organisations to execute mission-critical workloads closer to their applications and operations for swift, high-performance data access while maintaining control over data locality. Furthermore, SQL Server on Azure Local Disconnected Operations (ALDO) is now also generally available, allowing for consistent platform performance and experience in environments without cloud connectivity, restricted, or purposely disconnected.

Whether catering to regulated industries, operating at the edge, or managing critical business systems in disconnected environments, clients can now leverage modern Azure and database capabilities closer to their data while retaining the performance, control, and adaptability their workloads demand.

Finally, Microsoft Fabric is extending to Microsoft 365 Government Community Cloud High (GCC High) customers, accommodating organisations that require United States sovereign cloud environments and compliance with regulatory standards.

To learn more about the breadth of innovations we’re introducing in our database offerings, read Shireesh Thota’s blog, SQLCon Barcelona 2026: Advancing SQL with Greater Control, Scale, and Intelligence.

Explore More Innovations in Fabric and Azure Database

AI transformation will be defined not only by the number of agents organisations develop but also by the efficacy with which these agents comprehend the business and collaborate to deliver results. The organisations that establish a trustworthy, interconnected data foundation today will be best positioned to thrive in the future.

For a closer look at the updates highlighted above and to uncover additional features launched this month, please visit the Fabric September 2026 Feature Summary blog, the Power BI September 2026 Feature Summary blog, and the latest updates on the Fabric Updates channel and the Azure Data Tech Community blog channel.

Join Us at Microsoft Ignite

Don’t miss Microsoft Ignite in San Francisco from November 17 to 20, where you can learn from experts, gain hands-on experience, and discover the latest trends in AI and cloud technology. Choose from sessions tailored to your interests, with new content added regularly. Register now.

Discover More Resources for Microsoft Fabric